未明确命名

本研究聚焦于通过合成数据集生成方法，以医疗干预室中的医疗服装检测为例，探讨如何减少真实数据的需求。研究使用了3D扫描和设计的服装，在域随机化和结构化域随机化场景中进行比较，同时结合混合现实数据，旨在缩小合成数据与真实数据之间的差距。数据集包含不同类型的医疗服装，如手术服、手套、口罩等，用于训练和验证医疗服装检测模型。该数据集的应用领域主要集中在医疗干预环境中，解决医疗服装的自动检测问题，以提高医疗操作的效率和安全性。

This study focuses on exploring how to reduce the demand for real data through synthetic dataset generation methods, taking medical apparel detection in medical intervention rooms as an example. The research uses 3D-scanned and designed garments, compares scenarios between domain randomization and structured domain randomization, and incorporates mixed reality data, aiming to narrow the gap between synthetic data and real-world data. The dataset covers various types of medical apparel such as surgical gowns, gloves, masks and more, and is utilized for training and validating medical apparel detection models. Its application fields mainly focus on medical intervention environments, addressing the automatic detection of medical apparel to improve the efficiency and safety of medical operations.